JP4334770B2 - Communication device and packet communication method with high reliability and short delay - Google Patents

Abstract

In the present invention, the implementation of a protocol that embeds packets is made sensitive to the type of data being embedded, so that adjustable parameters can be set automatically at the layer providing the parameters being adjusted in accordance with the packets being embedded. Thus it is taken into account that the data in packets being received from a higher layer in order to be encapsulated or segmented and sent further (either directly into the link or on to lower layer), may be classified into different categories with respect to the adjustable parameters, where the parameters can automatically be adjusted for an individual packet in accordance with the category into which is classified. <IMAGE>

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a communication device and method for sending data packets over a connection established between a computer and a packet switched network, for example called the Internet.
[0002]
Background of the Invention
With recent advances in data processing and communications technology, it is essential to establish a data exchange network in many aspects of modern society, such as the scientific and military world or rapidly developing business and personal use. . The best example of this is the provision of services and the demand for services that are making rapid progress in a worldwide communication network called the Internet. Basically, the term “Internet” is a very well-known term, but this term simply refers to a network made up of other networks, and the so-called world-wide “Internet” It is only an example.
[0003]
The basis of data exchange widely used in networks is packet exchange. That is, the data to be transmitted is divided into units. The unit transmission / reception rules and the unit configuration rules are defined by a so-called protocol. A protocol is a set of rules that enables communication between the sender and receiver, and rules allow the receiver to interpret and respond to data under a protocol that binds the sender and receiver. In order to do so, it defines how the transmission data should be structured and sent.
[0004]
An important concept that enables data exchange between different networks, that is, between many software and hardware, is protocol layering. This is because many protocols (sometimes called sweet) have a layered layer structure, and each layer has its own functions and responsibilities. This point will be briefly described with reference to FIGS. A more detailed explanation of this concept can be found in It is described in Richard Stevens, “Illustration TCP / IP, Volume 1, Protocol”, 1994, published by Addison Wesley.
[0005]
The communication control protocol (TCP) and the Internet protocol (IP) are examples of protocols that can communicate on the Internet. However, what is usually called TCP / IP includes other protocols. The so-called TCP / IP sweet has four layers as shown in FIG. The link layer handles input / output of data to / from physical communication media such as cable and wireless communication. In a typical case, the link layer is incorporated in a device driver in a network interface corresponding to the OS. The network layer moves or routes packets within the network. IP is an example of a network layer protocol. The transport layer supplies the data flow to the upper application layer. TCP is an example of a transport layer. An alternative to TCP is the User Datagram Protocol (UDP). The application layer includes a program for transmitting / receiving data, that is, an e-mail program.
[0006]
An outline of mutual communication between two points is shown in FIG. Data is first dropped down through the hierarchy and input to the physical connection on the sending side, and then moves to the information on the receiving side. This movement is indicated by solid arrows. An element that performs data processing in accordance with a protocol given to the layer in a predetermined layer (TCP in the transport layer shown in FIG. 4) is also called a peer. In other words, the transport layer illustrated in FIG. 4 is processed by the PCT peer. It should be noted that the physical connection shown at the bottom of FIG. 4 can be one or more protocols in the layer depending on the type of connection. This point will be described in more detail below.
[0007]
FIG. 5 shows data transport between layers. A typical basic structure of a packet consists of two parts: a header and a payload part. The header has information related to communication control of the layer corresponding to the type of data included in the payload portion. As shown in FIG. 5, a header is added to user data in the application layer. The resulting data unit or packet is sent to the transport layer where a TCP header is added. This TCP header contains information for the receiving TCP peer. The resulting packet is sometimes called a TCP segment. This TCP segment is sent to the network layer, and according to the example shown in FIG. 5, an IP header including routing information is added. The resulting packet is sometimes called an IP datagram. Finally, the IP datagram is sent to the link layer, and a header for the link layer protocol (LLP), ie, point-to-point protocol (PPP) is added. The resulting packet is commonly referred to as a frame. A start flag and an end flag are added to the frame so that the reception side can recognize the start and end of the frame. A procedure for embedding such an upper layer packet in a lower layer packet is referred to as encapsulation.
[0008]
On the receiving side, user data is retrieved by performing de-encapsulation in the reverse order as shown in FIG. It should be noted that the above description is an example, and many variations are possible. For example, some protocols add tailors as well as headers.
[0009]
Note that the packet does not have to depart from the application layer, and that much lower layers may create the packet to deliver information about that layer only to the corresponding peer at the receiving end. It is. As an example of this, link control protocol (LCP) packets can be sent before sending higher layer data, and the LCP packets can be adjusted and / or negotiated by link layer peers (physically established ) To form a link.
[0010]
FIG. 4 illustrates the concept of four peers on each side of transmission and reception. The real situation is more complicated. FIG. 6 shows the concept of communication between the IP-host 1 and the IP-host 2. The top part of the figure shows physical communication paths, and the lower part of the figure shows logical communication paths corresponding to various protocols. The example shown in FIG. 6 relates to an Internet connection via a GSM network.
[0011]
The IP-host 1 may be a laptop computer. The terminal application function TAF is performed by the PCMCIA interface card connected to the mobile station MS, that is, the mobile phone. The mobile station MS communicates with the radio base station BTS connected to the base station controller BSC. The connection to the public switched telephone network PSTN is connected to the public switched telephone network PSTN by a modem in the interworking function IWF of the mobile switching center MSC corresponding to the radio base station BTS. Note that because GSM is a digital line, TAF is not a modem. Eventually, the PSTN provides a connection to the Internet service provider ISP, which becomes the gateway to the Internet. The physical connection between the Internet and IP-host 2 which is a personal computer to which e-mail from IP-host 1 is sent is not shown.
[0012]
As shown in the lower part of FIG. 6, a TCP peer, an IP peer, and a PPP peer are embedded in the IP-host 1. Corresponding TCP peers and IP peers are provided in the IP-host 2, and IP peers and PPP peers are provided in the Internet service provider ISP.
[0013]
One important aspect of the stratification scheme is that the different hierarchies are “transparent”. This means that one layer's peers are self-evident to occur in another layer. For example, the PPP peer establishes a pseudo PPP connection between the IP-host 1 and the Internet service provider ISP. The two PPP peers move independently of the type of physical connection being used, i.e., independent of which protocol is used in the lower layers. Similarly, PPP peers pass packets coming from higher layers transparently, i.e. without sensing their contents.
[0014]
In the example shown in FIG. 6, a forward error correction (FEC) layer and an interleave layer are provided between the mobile station MS and the radio base station BTS, and a connection by layer 2 transmission (L2R) and radio connection protocol (RLP) is provided. Is established between the terminal adaptation function TAF and the mobile switching center MSC; 42 and V.I. 32 interfaces are established between the intermediate operation function IWF and the Internet provider ISP. Since these protocols are all known except for RLP, no further detailed description will be given below.
[0015]
The RLP is responsible for moving data from the upper layer, but does not do this by encapsulation, that is, the upper layer data unit is not a single lower layer data unit. It should be noted that the RLP performs the division, that is, the upper one data unit is divided into a plurality of segments and each segment is inserted into one lower data unit. In other words, the RLP data unit is smaller than the upper layer data unit that it handles (ie, a PPP frame). The basic idea of inserting a higher-layer data unit into a lower-layer data unit is referred to as embedding in this specification, and is a concept that encompasses encapsulation and division.
[0016]
As will be understood from the above description, packets corresponding to different protocols are sometimes referred to by different names such as frames, segments, datagrams, etc. For clarity, the term “packet” is used herein generically to mean a data unit having a defined syntax and having a data structure predetermined by the protocol. “Data structure” means that specific data at a specified position (address) has a specific meaning according to a corresponding protocol. An example of this will be described with reference to FIG. 7 which shows a frame structure based on the high level data link control protocol (HDLC). HDLC is a very basic protocol, and PPP employs many of the basic characteristics of HDLC.
[0017]
FIG. 7 shows an HDLC frame having an address field, a control field, an information field, and a check field. In the HDLC, the address field, the control field, and the information field are variable lengths, but the check field has a fixed length determined by the protocol. For example, the information field is simply 8 characters following the control field and before the check value.
[0018]
Thus, a field is an example of a position defined in the above sense, and the protocol specifies that a particular number within a particular field should be interpreted according to a predetermined method, ie its value is determined by the protocol. Has a specific meaning.
[0019]
Problems to be solved by the present invention
It is known to assign a communication reliability mode when sending a packet. That is, the above-mentioned HDLC has two communication reliability, that is, a so-called numerical mode (or I-mode) and a so-called unnumbered (UI-mode) mode.
[0020]
In numbered mode, if the sent packet is not received correctly at the receiving peer, the sending peer will send the packet again. Based on the number of packets transmitted, in this way, all packets are delivered correctly, albeit with a delay. Information that the packet was received correctly is generally conveyed by a confirmation message from the receiving peer to the sending peer, and the exact flow control of transmission and retransmission is controlled by the moving window (so-called window-based flow control), This is done by transmitting a plurality of packets before receiving a confirmation, that is, without confirming the reception of each packet every time a packet is transmitted. This flow control is well known in the art, for example from the above-mentioned Stevens book and will not be repeated here.
[0021]
Unlike the numbered mode, retransmission is not performed in the unnumbered mode. This means that the packets are only transmitted in a predetermined order, regardless of whether the packets were received correctly. This mode is characterized by fast transmission, but the reliability of transmission depends on the quality of the physical connection.
[0022]
The reliability of transmission becomes a problem particularly in the case of wireless connection. In the example shown in FIG. 6, the radio connection protocol RLP is used. All known settings of RLP are fixed to use numbered mode. In this way, reliable data transmission is ensured by wireless connection whose communication quality varies greatly. As shown in FIG. 6, a system (referred to as a transparent circuit switching GSM data service in the example shown in FIG. 6) is configured so that the mobile station MS can request content to perform RLP in connection setup. It is known to compose. This can reduce the transmission delay, but instead reduces the transmission reliability.
[0023]
The problem is not limited to RLP or wireless connections, but exists for any protocol that provides a reliable mode, such as the numbered mode described above.
[0024]
Furthermore, in general, there is a problem as to which mode should be selected in order to transmit a packet in a specific layer.
[0025]
Comments from Blake et al. (RfC) 2475, December 1998 (http://www.ietf.org/html.charters/diffserv-charger.html), architectures for different services are known. RfC 2475 proposes to classify and index packets so that they can receive forward behavior (PHB) for each specific hop at nodes present in the path. The header information of the header of a packet belonging to a predetermined protocol layer is used only for the operation of that protocol layer. The differentiated service architecture is based on a model in which traffic entering the network is classified at the boundaries of the network and in some cases conditional, and associated with different sets of behavior. Each set of behavior is distinguished by a single differentiated service (DS) code point. A packet classification function for selecting packets in a traffic stream based on the contents included in a part of the packet header is described. As described above, the DS code point and DS field belonging to the IP layer are used only in the IP layer. Behavior set classification means for classifying packets based only on DS code points is described. Similarly, multi-field classification means for classifying a packet based on a value composed of a combination of one or more header fields is described.
International Publication No. WO95 31060A discloses a method for transmitting a data packet based on the type of message. When data is transmitted from the information source to the terminal, the network interface can select a packet switch network or a circuit switch network based on the type of message sent. The purpose of the disclosed method is to properly route data to use the best path through either a packet switch network or a circuit switch network. For example, short messages such as email need to be sent through a packet switch network, while long messages such as file transfers need to be sent through a circuit switch network.
[0026]
Object of the present invention
It is an object of the present invention to provide a mechanism that solves the above-described problem and realizes a protocol that can perform mode selection control more efficiently, that is, improve the relationship between transmission reliability and transmission speed. It is.
[0027]
DESCRIPTION OF THE INVENTION
The above object is directed to a data packet having a first data structure defined by a first predetermined protocol, configured to receive a data packet having a second structure defined by a predetermined second protocol. And creating a data packet having the first structure by embedding one or more data packets having the second structure in one or more data packets having the first structure. This is achieved by an apparatus as set forth in the independent claim having discriminating means configured to distinguish the data packet having the second structure according to a predetermined rule based on the content of the data packet having the structure.
[0028]
Preferred embodiments of the invention are described in the dependent claims.
[0029]
According to the present invention, the execution of the protocol depends on the type of data to be transmitted, and the setting of parameters related to transmission can be performed in a layer that automatically adjusts according to the content of the transmitted packet. In other words, the present invention encapsulates, segments, and forwards (whether directly to the link or through lower layers, taking into account that data is received from higher layers). Regardless, for adjustable parameters, the parameters are classified into different categories that are automatically adjusted for each packet according to the classification.
[0030]
The term content is used in a broad sense including any content. The specific type of content used for discrimination may be added with care to the data at the time of packet production at a specific layer, taking into account the lower layer classification operations, ie A field can be used to identify information for a lower layer peer to a lower layer peer so that a higher layer peer affects processing in the lower layer, or The discrimination process in the lower layer is independent and individually by analyzing the protocol ID of the header that is not intended to convey information to the lower layer without the influence of the upper layer, that is, the header It should be noted that it can also be done.
[0031]
The present invention is therefore essentially inconsistent with the firmly established criteria that all layers must be used independently of each other. In contrast to what has been known so far, the present invention provides for the possibility of setting certain parameters (ie reliability) in the upper layer in the lower layer when establishing a connection. Information is given to a predetermined protocol in each layer, and parameters are set flexibly and according to what type of packets are continuously transmitted according to its own decision.
[0032]
According to a preferred embodiment of the present invention, the adjustable parameter is related to the reliability mode, and the reliability mode (numbered or unnumbered mode) is automatically determined in the embedding layer according to the determination result. Adjusted.
[0033]
According to this preferred embodiment, there is an advantage that data packets having different transmission reliability modes to which the present invention is applied can be automatically and simultaneously transmitted between a plurality of peers. For example, if the present invention is used for embedding PPP packets by segmentation and encapsulating higher layer packets by IP or RLP in higher TCP or UDP, UDP can be performed by executing RLP according to the present invention. A packet including a packet can be distinguished from a packet including a TCP packet. Therefore, in the RLP layer, a UDP packet can be transmitted in an unnumbered mode and a TCP packet can be transmitted in a numbered mode.
[0034]
As already mentioned in this example, the embodiment relating to the setting of the reliability mode is preferably applied to the implementation of a link layer protocol (eg RLP). Similarly, classification into categories is preferably done taking into account the reliability rules of the sorted higher layer packets. In the above example, UDP packets are classified into the category transmitted in the unnumbered mode, which is an advantage considering that UDP does not retransmit packets that contain errors. In contrast, TCP offers the possibility of retransmission. By applying the present invention, it is possible to simultaneously transmit a TCP packet and a UDP packet in the RLP layer. Despite this method, the communication reliability mode of each packet in the RLP layer is a numbered TCP packet. Transmit under the desired reliability of the required mode and no unnecessary delay due to retransmission while transmitting UDP packets in the same numbered mode.
[0035]
With respect to the above example, the invention has significant advantages. PPP is used to simultaneously transmit a plurality of protocols such as IP, LCP (Link Control Protocol), PAP (Password Authentication Protocol), etc. via a serial link. The PPP protocol identifier indicates which protocol data unit (PDU) or protocol packet is included in a particular PPP packet. The IP sent by PPP can send TCP, UDP and ICMP (Internet Control Message Protocol) multiple protocol packets, which are also distinguished by the protocol identification in the IP header. Protocols transmitted by PPP differ in trade-off requirements regarding reliability and delay. Therefore, it is not uniform whether the numbered mode or the unnumbered mode is preferred for a particular data stream. More specifically, signaling messages such as LCP and PAP must be sent in numbered mode, and real-time data streams (eg, voice signals for internet telephone conversations sent in UDP packets) usually result in data loss. Since it is strong but sensitive to delay, transmission in the unnumbered mode is preferable.
[0036]
Previously known implementation methods and systems cannot solve this problem. If the RLP is operating in a fixed numbered mode, a real-time UDP data stream cannot be sent simultaneously on the mobile computing device and a data stream that requires the numbered mode (eg, TCP). If the RLP is not used, the PPP connection may be performed after the circuit switching connection is achieved, that is, after the charging is started, because the data of the LCP packet may change in the unnumbered mode. Can not. Therefore, the user cannot receive a necessary service despite having to pay a usage fee.
[0037]
The present invention does not introduce any delay by the execution of RLP automatically sending LCP messages in numbered mode to achieve a secure PPP link, followed by sending PPP packets including UDP packets in unnumbered mode. This solves this problem. In addition to solving this problem of sending different types of packets at different times on the same connection, the present invention simultaneously transmits two (in terms of reliability) opposite streams in parallel, eg the UDP described above It is possible to transmit a TCP packet simultaneously with a data stream.
[0038]
It should be noted that the above protocol and the above state are only examples used to illustrate embodiments and advantages of the present invention. As already mentioned, the embodiment for setting the reliability can be used to implement a protocol that allows more than one reliability mode. For example, the present invention can also be used when an LLC (Logical Link Control) packet is transmitted over an RLC (Radio Link Control) link. Similarly, the implementation location is not limited to the example shown in FIG. 6 above, and can be used anywhere that is appropriate or necessary in a given protocol. For example, in general packet radio service (GPRS), they are TAF and GSN (GPRS support node).
[0039]
As a result, this means that by applying the present invention to the reliability mode setting, a specific packet is automatically transmitted to a specific virtual channel corresponding to the number of communication reliability modes provided or selected. Means that a virtual communication reliability channel is provided. In the above example, numbered and unnumbered channels are provided, so that TCP packets (more precisely, packets with embedded TCP packets) are sent on numbered channels and UDP packets are sent on unnumbered channels, Since two transmission reliability modes are used, two virtual channels will be provided.
[0040]
Expressing the above example more generally, the present invention describes an ARQ-compliant error correction (ARQ is an ARQ) that can be based on the contents of the packet in which the choice of operating mode (in this sense of ARQ, reliability mode) is embedded. (Automatic repeat request). Of course, the applicable modes are not limited to two modes, that is, untrusted (UI mode) and trusted (I mode). In addition to untrusted (UI mode) and trusted (I mode), the above two modes are applicable. In addition, the number may be larger, such as having a semi-reliable mode with delivery in sequence or delivery out of sequence.
[0041]
As described above, the present invention can be provided for setting the reliability mode, but is not limited thereto. Quite the contrary, the present invention can be applied to setting or adjusting all kinds of processing modes related to embedding higher layer packets in lower layers. More specifically, the present invention can be applied to all communication schemes having at least two modes. For example, the receiving side discards frames with errors, the receiving side marks frames with errors but does not discard them, or the receiving side does not mark frames with errors and does not discard them. 2 is an FCS (frame check sequence) error detection communication scheme having an operation mode that is based on two modes or all three modes. Alternatively, the present invention can be applied to FEC-compliant error correction (forward error control) using an operating mode using two or more FEC schemes. Alternatively, the present invention can be applied to an error prevention communication scheme based on interleaving using an operation mode using two or more interleaving schemes, for example. Alternatively, the present invention can be applied to, for example, a power control communication scheme using an operation mode using two or more power levels. Furthermore, the present invention can be applied, for example, to communication schemes based on spreading using two or more spreading factors (eg spread spectrum techniques such as CDMA). In addition, the present invention can be applied to a frame length control communication scheme using an operation mode that can take two or more frame lengths, for example. Further, the present invention can be provided for a bandwidth conservation control communication scheme using an appropriate operation mode such as 1, 2, 3,..., 8 times slot allocation such as GPRS communication standard.
[0042]
One of the important advantages of the present invention is that it basically does not require any protocol changes and can be broadly adapted to any content without any compatibility or migration issues due to only peer implementation changes. It is an applicable point.
[0043]
According to another preferred embodiment of the present invention, the present invention provides a virtual network according to the communication quality requirements specified by the upper layer peer when the upper layer packet or data unit is embedded in the lower layer. Packet communication can be performed at the layer (eg, QoS (Quality of Service) requirements and / or communication characteristics). There is a standardized scheme in which upper layer peers use the layer packets as vehicles for sending quality requests to lower layers, ie the quality requirements are understood in both layers. For example, this is done by using a predetermined field in the header of the upper layer packet and standardizing the content contained in that field. In other words, some kind of control channel is provided to communicate quality requirements from higher layers to lower layers. This quality request not only conveys upper layer packets and flows (the flow is a continuous packet that can be specified) to the lower layer for operating mode, but at least in part based on the quality requirements. For, dynamically apply the operation mode by itself.
[0044]
The communication quality request is control information that can be designated and used as a quality command that causes the upper layer to set a specific operating mode for the lower layer, or the lower layer carries the quality request. It should be noted that the operation mode is freely selected according to the quality requirement for a specific packet, but the upper layer does not directly control which mode is selected by the lower layer. .
[0045]
Other aspects and advantages of the invention will become more apparent from the following detailed description of preferred embodiments of the invention with reference to the drawings, which illustrate preferred embodiments of the invention.
[0046]
The present invention will be described with reference to an embodiment relating to setting of the communication reliability mode. As already mentioned, this is a preferred embodiment of the present invention and is described for the purpose of illustrating the present invention. However, the invention is applicable to setting all kinds of adjustable parameters and can also be used for setting the security mode.
[0047]
The basic configuration of the present invention will be described with reference to FIG. In this embodiment, a communication device equipped with a layer n communication protocol receives a packet from the upper layer (n + 1) and embeds the packet in a layer n packet. As already mentioned, the term “embed” is a concept that includes both encapsulation and segmentation. It should be noted that all descriptions of the introductory part of the present invention are incorporated into the description of the present invention to avoid duplication of explanation.
[0048]
For example, the layer n protocol may be RLP and the layer (n + 1) packet may be a PPP frame, in which case embedding means dividing the PPP frame or packet into a plurality of RLP packets. Become. PPP frames typically have a length of 1500 bytes, whereas RLP frames typically only have a length of 30 bytes. As another example, the layer n protocol is PPP, and the layer (n + 1) protocol is a higher-level appropriate protocol, such as the protocol shown in FIG. This is done by encapsulation.
[0049]
According to the invention, the embodiment comprises a discriminator 1 which receives a layer n + 1 packet and discriminates the packet received by this discriminator according to a predetermined class classification rule. In other words, this means that the discriminating device confirms a specific feature of the received packet and classifies the received packet based on the feature. As one example of this, the packet of layer n is a PPP frame, the individual header of the received PPP frame is checked to determine what kind of information is sent by this frame, and the frame is determined according to the result. Classify. For example, if a frame carries LCP information, it is classified as “reliable transmission”, and if it carries UDP information, it is classified as “no reliability required” category. .
[0050]
The above protocols and categories are exemplary only, and the invention relates to all types of discriminators that can classify received packets according to their contents. Therefore, any protocol can be used for these packets, and the number of categories in which the packets are classified is arbitrary. The number of categories may be the same as the number of communication reliability modes provided by the layer n protocol, but may be less if the layer n protocol provides more than one mode. .
[0051]
As a result of the determination, that is, the class classification is transmitted to the control means 2, and the control means 2 next controls the embedding means 3 by a communication reliability setting command. In other words, the discriminator uses the embedding means 3 as a packet to be embedded, where it is embedded in one or more packets of layer n, and communication reliability appropriately set by the corresponding flag for each packet of layer n The security mode is set according to the communication reliability setting command.
[0052]
Next, the layer n packet created in this way is transmitted to the next processing stage. The next processing stage varies depending on the application, and can be appropriately selected according to a desired or necessary method. For example, if a layer n packet is intended to go directly to the next (not shown) virtual layer (n−1), the layer n packet is simply passed to the next layer input buffer. Also good. However, if it is envisaged that packets will be sent out through the established link, they are preferably placed in the output buffer and appropriately controlled according to the situation in question. The appropriate control of data transmission from such an output buffer depends on the application, ie the layer n protocol and its link.
[0053]
As can be seen from the above description, the present invention can be applied to any protocol that embeds one layer packet in another layer packet and provides at least two different reliability modes. It can be very flexible.
[0054]
FIG. 2 is a flowchart illustrating a method according to the present invention. In the first step S1, it is determined whether or not a higher layer packet has been received. When an upper layer packet is received, the process proceeds to S2, and the received packet is determined according to a predetermined class classification rule. Next, in step S3, the received packet is embedded in one or more packets of a predetermined layer (n in FIG. 1), and the communication reliability mode of each packet is set according to the determination result. Finally, the embedded packet is sent to the next stage, where it is sent to the output buffer for transmission in step S4 by the process shown in FIG.
[0055]
FIG. 8 shows a system corresponding to the system shown in FIG. 1 and having an input / output buffer. As shown in the figure, the apparatus shown in FIG. 8 also has a discriminator 1, a control means 2 and an embedding means 3, and an input buffer 4 and an output buffer 5 are shown. Both are connected to the control means 2 and controlled thereby. This operation is the same as that described with reference to FIG. 1, except that the layer n packet is supplied to the discriminating device 1 through the input buffer 4, where the control means 2 causes the input buffer 4 to pass the packet to the discriminating device 1 as appropriate. 8, the embedding unit 3 sends the layer n packet to the output buffer, and then the layer n packet is sent out under the control of the control unit 2, as exemplified in FIG. 8. Is different.
[0056]
As already described, the present invention has the flexibility of being able to select any suitable flow control. For example, if there are two communication reliability modes such as the numbered mode and the unnumbered mode described above, the corresponding packets in layer n can be placed in separate queues in the output buffer 5. In other words, if there are two communication reliability modes that can be set for the layer n packet, the output buffer has two queues, one for numbered mode packets and the other for unnumbered packets. It will be.
[0057]
In general, in one embodiment of this system shown in FIG. 8, the output buffer 5 has a queue corresponding to each communication reliability mode, and the packet is discriminated by the discriminator 1 and each of them is discriminated. It is taken to the queue. Packets in each queue can be handled in any suitable or desired manner, i.e., each queue can have a preset priority (inherent reliability mode). In the case of having only numbered mode and unnumbered mode, there is no basic priority between the numbered packet and the unnumbered packet with respect to the packet to be sent first (in which queue these packets are The numbered packets that must be retransmitted have a high priority and are therefore transmitted preferentially. In such a case, control of the flow from the numbered mode queue is done by simply sending these packets in the proper order.
[0058]
However, as will be described below with reference to other embodiments, a method different from the above can be adopted.
[0059]
A preferred embodiment of the present invention is shown in FIGS. 9a and 9b. Steps already described in connection with FIG. 2 have the same reference numbers. The method illustrated in FIGS. 9a and 9b corresponds to the case where a protocol is provided that provides a numbered reliability mode (I-mode) and an unnumbered reliability mode (UI-mode). In other words, in numbered mode, the confirmation package informs the sending peer that it has received it correctly, and if the packet is not received correctly by the receiving peer, the packet is retransmitted. Will not retransmit.
[0060]
In FIG. 9, steps that are the same as or correspond to those shown in FIG. 2 are given the same numbers. In the first step S1, it is determined whether or not an upper layer packet has been received. If so, the process proceeds to step S21, and the protocol ID included in the upper layer packet received there is analyzed. For example, if the upper packet received is a PPP frame as shown at the bottom of FIG. 5, the process simply checks the PPP header directly following the start flag. Next, in step S22, it is determined whether or not the confirmed protocol ID is in the numbered mode, and this classification is performed according to the above-mentioned preset rules. In the case of a PPP packet, it is preferable that the above-mentioned rule is to put the LCP packet into a numbered mode. This is because LCP packets are part of a link establishment procedure that requires a high level of reliability. Therefore, if it is determined in step S22 that the LCP information has been sent as a result of the process, the process branches to step S31, and the upper layer packet received there is the layer in which the present invention is being executed. Embedded in one or more packets, the reliability mode of each embedded packet is set to a numbered mode by setting the corresponding reliability mode flag for each packet.
[0061]
If it is confirmed that the protocol ID specified in step S22 is not in the numbered mode, it is determined in step S23 whether the protocol ID is in the unnumbered mode. If so, the process proceeds to step S32, where the received higher layer packet is embedded in one or more packets of the layer in which the present invention is implemented, said one or two. Each reliability mode of one or more packets is set to an unnumbered mode by setting an appropriate flag. If the results of steps S22 and S23 are both negative, step S24 indicates that the check is made up to the protocol layer of the received packet, and the next header is checked. Returning to the example shown at the bottom of FIG. 5, this is an IP header that includes a protocol ID for identifying TCP, UDP or ICMP data. Next, steps S21 to S24 are repeated. If the identified protocol ID cannot be identified again as the numbered mode or the unnumbered mode, the process goes up again to the protocol layer, in the case of the example shown in FIG. 5, the TCP header.
[0062]
In general, the relationship between a predetermined protocol packet and a numbered mode or an unnumbered mode is stored in a table referred to from determination steps S22 and S23. Ideally, all possible protocols are recorded in the table, and all received packets should be classified as numbered or unnumbered mode. However, from an economic point of view, the process shown in FIGS. 9a and 9b is provided with a fail-safe system, i.e., a counter with a limited number of times that it can go to the next protocol layer in step S24. It is preferable that a default mode is set for a received packet when a predetermined number of times is exceeded. This default mode may be a numbered mode or an unnumbered mode. Of course, other fail-safe structures are possible and can be selected as needed or desired.
[0063]
Returning to FIG. 9b, after steps S31 and S32, the process proceeds to step S4, where the packet is sent to the output buffer, similar to that shown in FIG. The process then returns to step S1 and waits for the next higher layer packet to be received.
[0064]
In general, according to the method shown in FIG. 9, a data packet has one or more sections including information on the contents of the packet (packet encapsulating a protocol hierarchy), and the discriminating means is the one described above. The above section is analyzed to determine the data packet according to its contents. More precisely, the one or more sections are packet headers corresponding to the corresponding protocol layer and including protocol identification information that can identify the protocol with which the packet content is associated. The packet header has a hierarchical structure depending on the protocol layer, and for a packet to be discriminated, the discriminator first determines the protocol identifier of the header corresponding to the protocol of the first higher layer, and then determines this protocol identifier The protocol identifier and a preset communication reliability mode are made to correspond with the stored rule. Further, the discriminating device sets the communication reliability mode of the packet based on a predetermined response if the communication reliability mode of the packet can be found in the rule in which the protocol identifier is stored. Is not found in the stored rule, the protocol identifier contained in the header associated with the next protocol one level higher in the hierarchy is read, and the next protocol identifier is stored in association with the communication reliability mode. Compare with the rules. Here, in the process of determining and comparing the protocol identifier, the protocol identifier determined to be discriminated in the packet is associated with one of the communication reliability modes determined in advance according to one of the rules. The security mode is set for the packet to be determined, or the default communication reliability mode is set for the packet to be determined when the fail-safe condition is satisfied.
[0065]
The present invention is a protocol for performing division when embedding upper layer packets in a packet to be transmitted via a link, and is applied to one that executes two kinds of communication reliability modes, numbered mode and unnumbered mode. Is preferred. An example of this type of protocol is a protocol that handles radio link communication such as RLP and RLC described above.
[0066]
The application of the present invention to this type of protocol is preferably performed according to the embodiment shown in FIGS.
[0067]
In such a case, the present invention is not only characterized with respect to the embedding of the packet to be transmitted, but also with respect to receiving such an embedded link layer protocol from the sending peer as an implementation as a receiving peer. It is preferred to have characteristics with respect to properties.
[0068]
In the following, some basic mechanisms of peers of the link layer protocol used as transmitter and receiver will be described using RLP as an example of a link layer protocol that divides and embeds PPP frames. . However, it should be noted that these protocols are merely illustrative and that the mechanisms described herein can be applied to any other protocol where appropriate.
[0069]
Based on the basic configuration, the RLP transmitter should analyze the data stream received from the upper layer, that is, the PPP frame, and in which mode, ie, numbered mode or unnumbered mode, each upper layer frame should be transmitted. To decide. Since this processing is executed by the method shown in FIGS. 8 and 9, it will not be repeated here. Flow control from the output buffer allows the first transmission of numbered mode and unnumbered mode RLP packets (for numbered mode packets because unnumbered mode is always sent once) to the FIFO. Therefore done. However, numbered mode RLP packets that must be retransmitted have a high priority and are therefore always sent as quickly as possible.
[0070]
The RLP receiver separates the unnumbered mode and numbered mode packets and puts them in their own separate buffer or at least another queue. Similar to the RLP transmitter, the RLP receiver always analyzes the data contained in each buffer and thereby discriminates higher layer frames. When a higher layer frame is received only once, it is sent to the higher layer on the receiving side. For packets sent in unnumbered mode, the higher layer packet delimiters (i.e. start or end flags) may not have been retransmitted and may have changed in the middle of the communication. May happen. In this case, the RLP receiver uses a fixed buffer limit for the unnumbered mode packet, and the unnumbered mode packet exceeding this is sent to an upper layer regardless of whether or not the delimiter is received. More precisely, when a fixed buffer limit is exceeded, the RLP receiver sends all received unordered mode RLP packets to the next higher layer until the next higher layer packet delimiter is detected. With the above upper layer packet delimiter as a new starting point, the RLP receiver can also wait for the next arrival of the upper layer packet delimiter, i.e. the next complete upper layer packet is received. I can wait.
[0071]
The method of distinguishing each upper layer packet by the RLP receiving apparatus is to copy all of the upper layer delimiters in the data transmitted by the RLP transmitting apparatus and transmit them. In other words, instead of discriminating the data according to its contents, the RLP transmitter also searches for the upper layer, that is, the PPP delimiter, and always inserts the same delimiter next to the detected delimiter. This method has the advantage when the fixed buffer limit that has been exceeded is within the RLP reception characteristics, that is, when a copy of the transmitter delimiter is combined with the buffer limit characteristics of the transmitter.
[0072]
Other mechanisms related to the RLP sender and receiver will be described. The transmitter features are the same as before, that is, the transmitter analyzes the received data stream to determine in which mode the higher layer frames are to be transmitted, ie numbered or unnumbered mode. Flow control from the output buffer is done in such a way that the retransmitted numbered mode packets always have the highest priority. As for the RLP reception characteristics, the receiver uses the received numbered mode packet and unnumbered packet in its own buffer or queue as in the above example. In this case, only the receiving device analyzes the numbered mode packet (PPP frame) in order to determine the upper layer packet. In general, the receiving device passes the numbered mode and unnumbered mode packets immediately to the upper layer under the normal constraint that the numbered modes must be passed in the correct order. In other words, this is the case for numbered packets that are being retransmitted in order to avoid having to store a certain number of packets to pass in the correct order to higher layers. Except for this, the receive buffer normally does not hold the received packet, but immediately delivers it next.
[0073]
This rule on how the receiver passes packets to higher layers at the receiver side distinguishes four states.
(a) Only an unnumbered mode packet is transmitted (for example, a state called a UDP session, where the transmitting side transmits only a UDP packet), and there is no problem at all.
(b) Only numbered mode packets are transmitted (for example, a state called a TCP session, where the transmission side transmits only TCP packets), and there is no problem at all.
(c) When a numbered mode packet is transmitted following an unnumbered mode packet (for example, a TCP session is performed following a UDP session), there is no problem at all.
(d) Numbered mode packet is followed by unnumbered mode packet (for example, TCP session is followed by UDP session, or UDP session and TCP session are sent in parallel), numbered A problem can occur if an unordered mode packet is received when a mode packet requires retransmission.
[0074]
In order to solve the problem described in (d), the following mechanism can be used. As long as the upper layer packet sent in numbered mode has not been completely received (a packet containing a PPP frame delimiter indicating the end of the TCP packet must be retransmitted), the unnumbered mode packet is temporarily saved. There must be. The buffer is not passed to the upper layer until the entire numbered mode packet is received and passed. After the numbered mode packet is completely received and passed to the next higher layer, the buffer containing the unnumbered mode packet is passed. After that, normal operation continues.
[0075]
In accordance with the preferred embodiment of the mechanism shown above, the receiver further has knowledge or information regarding which packet delimiter is currently in use. The default delimiter is the normal packet delimiter one layer above (ie the PPP frame delimiter), but between the two link layers, ie between the two RLP peers, if other delimiters must be used. Can be negotiated. In other words, this means that upper layer protocol packets can be processed and one or more protocols can be processed. In the above example where the PPP frame is embedded in the RLP packet, this means that the execution of RLP not only embeds PPP, but also embeds frames of other protocols belonging to the same layer as PPP. . There are various ways in which the receiver can receive this information. In the case of a link layer protocol based on DHLC (for example, GSM or RLP) using a special signal frame or XID frame, which is a standard HDLC frame for determining link parameters between each other, the XID frame having a new parameter is used. be able to.
[0076]
Another mechanism related only to transmission characteristics will be described. In other words, there is no change in the receiver. Therefore, the receiver always sends out the unnumbered mode packet when it receives the unnumbered mode packet, and the numbered packets are sent out in the usual order. Therefore, it is the transmitter that ensures that higher layer packets (ie, PPP frames) are correctly passed on the receiver side. This means that before the transmitter starts an unnumbered mode session, the unnumbered mode packet must be completely terminated, that is, the numbered mode packet must be recognized by the receiver. is doing. This mechanism is easy to introduce because the receiving mechanism is not changed, but a delay occurs when the unnumbered mode session and the numbered mode session are switched.
[0077]
Finally, another mechanism is to keep the transmitter's output buffer, and when the transmitter is in an unnumbered mode session, all unnumbered mode packets are continuously sent in RLP frames with a flag indicating the unnumbered mode. It is a method of controlling to transmit.
[0078]
As already mentioned, the above example illustrates the use of RLP and PPP as preferred embodiments, but this mechanism is not limited to these protocols. Rather, they can be used for any protocol with appropriate characteristics.
[0079]
In the above-described embodiment, determination is performed using a specific field of a packet header of an embedded packet, that is, an ID field of an upper layer, and two operation modes related to reliability in a numbered mode and an unnumbered mode The invention has been described based on a specific communication scheme with ARQ-compliant error recovery using. However, the present invention is not limited to these in any sense.
[0080]
The present invention can provide a method for performing discrimination without limitation on where a packet is, for example, the DS field is not limited to the IP option or the like, and is not limited to the content of the packet. With respect to the operation modes associated with ARQ, more than two modes can be used, such as using a mixed mode in addition to the number mode and unnumbered mode already described.
[0081]
The present invention is not limited with respect to the communication scheme of ARQ, that is, error recovery, and can be applied to an operation mode using any form of forward error correction. More particularly, the present invention provides error detection based on at least two different modes of operation, eg FCS- (FCS means frame check sequence), ie the receiver discards frames with errors or the receiver detects errors. It can be applied to all modes of operation that distinguish between the two modes or the three modes. it can. The present invention can also be applied to FEC-based (forward error correction) error correction communication schemes using two or more FEC schemes. Alternatively, the present invention can be applied to an error prevention communication scheme based on interleaving using an operation mode using two or more interleaving schemes. Furthermore, the present invention can also be used in a power control communication scheme in an operating mode using more than one power level. In addition, the present invention can be applied to spread-based error prevention (eg spread spectrum techniques such as CDMA) communication schemes using modes of operation using more than one spreading factor. Furthermore, the present invention can be applied to a frame length control communication scheme using an operation mode in which two or more lengths can be selected as the frame length. Furthermore, the present invention relates to 1, 2, 3,. . . Alternatively, the present invention can be applied to a bandwidth saving control communication scheme using an operation mode such as 8-hour slot allocation, that is, a GPRS communication standard.
[0082]
Of course, any combination of these communication schemes and modes of operation is possible. For example, numbered and unnumbered modes and four power levels are eight modes (I-mode, power level 1), (I-mode, power level 2),. . . , (UI-mode, power level 4) means that discrimination can be performed in a general operation mode composed of a combination of a power control mode and an ARQ-related mode. As described above, any combination is possible as long as it is suitable for a specific application.
[0083]
Next, still another preferred embodiment will be described with reference to FIG. In the case of this preferred embodiment, the present invention is a communication in which, for example, a QoS (Quality of Service) request and / or a higher layer packet or data unit is embedded in a lower layer as specified by a higher layer peer. It is applied to a packet communication method in a lower layer according to communication quality requirements such as characteristics.
[0084]
In the following description, the term L3 is used to refer to a network layer protocol such as IPv4 or IPv6. See FIG. A data unit (PDU) according to the L3 protocol is called a packet, and the PDU header is called a packet header. The term L1 / L2 refers to one or more protocols below the L3 layer, for example the physical and / or link layers of a particular network. The term bearer is a quality of service (QoS) setting in which all L1 / L2 operation modes are fixed, and L1 / L2 (for example, wireless) with all L1 / L2 operation modes in between. It is applied dynamically depending on the conditions for each time. The term peer, when used in connection with a process, means a process that communicates with a common protocol at a particular layer. As an example, there is a server that communicates with the WWW client and the application layer using the HTTP protocol, and a TCP transmitter and receiver on the transport layer, see FIG. In the following context, the term peer always means a peer that communicates with a protocol based directly or indirectly on L3. The term flow means communication based on L3 packets from peer to peer (packet switching), identified by source address, source port, destination address-destination port and protocol ID. In other words, the flow is considered to have a specific QoS requirement. QoS requirements may change over time.
[0085]
It should be noted that if a peer (eg, a legacy WWW browser) has selected a QoS request for a flow, this involves having another process do so for the peer at the same time.
[0086]
FIG. 10 shows a mobile IP-host suitably linked to a mobile telephone adapted to IP packets (a so-called Bluetooth link is shown in FIG. 10). The phone can establish a wireless link through the radio access network and can use any communication standard such as GSM, GPRS or W-CDMA. Although only two radio access networks are shown in FIG. 10, they are connected to the Internet to which a large number of hosts are connected.
[0087]
The communication status is shown in the lower part of the figure. Peer A. 1 and B.I. 1 is shown as an example of a mobile station. 2 and B.B. 2 is a corresponding counterpart. These peers pass QoS requirements and / or communication characteristics, generally quality requirements as control information, to the L3 layer. Peers are L3 peers or higher layer protocols than L3. This identifies certain fields in the L3 header as being related to quality requirements, and the L3 layer allows the peer to use its QoS request and / or priority as a per-flow or per-packet level. Is done.
[0088]
The determination of the L3 packet is performed by the control device or control process 11 on each side of L1 / L2, and the control device or the like controls the adapter or adapter process 12 according to the determination result. The control process 11 reads the QoS requirements and / or characteristics for each packet and maps them to the corresponding L1 / L2 mode of operation and / or network characteristics. The apply process 12 can adapt the L1 / L2 mode of operation on a packet-by-packet basis and make this determination based at least in part on information provided by the controller. In a typical case, the application of the L1 / L2 mode of operation is performed based on the measurement status of the media existing under L1 / L2.
[0089]
As a result, the present invention has a dynamically adaptable, per-packet control process that can access the QoS requirements and / or characteristics of the flow / packet, which provides the providing process and / or L1 / L1, respectively. You will have L1 / L2 that can be used to control the L2 scheduler.
[0090]
As an example, the optimal application of L1 / L2 to TCP uses ARQ-based error recovery in I-mode or other communication schemes, eg FEC to optimize for the highest throughput. By reading the protocol ID field (indicating that it is a TCP flow), the control device can make a reasonable determination that the flow is, for example, a mobile terminal TCP flow. Optimal application for voice communication flow over real-time IP (wireless broadcast without session initiation protocol) has no error recovery based on ARQ (UI mode) or any other communication scheme, ie FEC is the source The transmission rate and the residual error are optimized to be the smallest. By reading the DS field of the packet header, the control device can make the optimum determination for the voice flow on the mobile terminal real-time IP.
[0091]
The above-described embodiments can solve the problems with the implementation of the known L1 / L2 protocol. L1 / L2 in the current network (wireless network) must simultaneously transmit different flows, each of which may have different QoS requirements. However, L1 / L2 on the current network cannot go down to a level that completely satisfies the QoS requirement of each flow, and is further lowered to satisfy the QoS requirement of each packet of the flow. You can't get down to the next level. The latter is necessary when the QoS request changes with time and the QoS of packets in the same flow are different. Instead, current networks can support at most one or more bearers simultaneously. However, the flow method is mapped to this bearer and is not directly related to the QoS requirements of these flows, but is based on criteria such as per-packet network operator charge characteristics (performed in GPRS communication systems). As if). The result is that the QoS requirements are not met for some of the flows / packets, or the QoS requirements are met for some flows / packets but not optimally (for TCP window flows). The FEC scheme selected for it is too strong compared to the lower bit rate provided by the optimal FEC scheme) or the flow / packet QoS requirements are overfilled and network resources are wasted That's it.
[0092]
The application of the L1 / L2 flow to the QoS request is independent of the priority-based traffic planning concept. For example, a real-time audio stream may have specific QoS requirements regardless of whether it is considered maximum effort or some sort of predictive QoS traffic class.
[0093]
This embodiment solves this problem by dynamically adapting L1 / L2 for each flow or QoS per packet. The method has the following functions and internal structures that realize them:
-Function 1: a standard scheme in which the QoS requirements for this flow / packet are specified exclusively so that the semantics are the same in any network that supports the scheme;
Function 2: An adaptation process that adapts L1 / L2 according to the operating mode supplied from the control process. The mode of operation may be provided explicitly or implicitly in a form that is local, that is, only meaningful for QoS specific to the L1 / L2 network. The adaptation process can finally determine the operation mode by combining the operation mode provided from the control process and the measurement information on the condition of the medium (wireless) located in the lower layer of L1 / L2. This is however only an implementation detail.
-Function 3: a control process that has access to the control channel and can control the adaptation process per flow / packet when a QoS request per flow / packet is given; and
Function 4: A control channel that makes per-flow / packet QoS requests available to the control process.
[0094]
The adaptation and control processes can each be a distributed process, one process per L1 / L2 communication scheme or a single adaptation / control process for all L1 / L2 communication schemes. This is also just an implementation detail. The functions described above need to be per packet or at least for packet sequences. That is, it is possible to change the FEC from one packet to another or from at least one packet sequence to another packet sequence.
[0095]
Regarding what is known from the current state of RfC2475, it should be noted that RfC does not suggest any adaptation of L1 / L2 dynamically. From the prior art, it is not known at all to adapt L1 / L2 to the QoS of this packet / flow (function 2), nor is it known to use a functional control process.
[0096]
Using a GSM-CSD or GPRS communication system as an example of a network belonging to the prior art, various scenarios can indicate that none of the networks has the capability to support flow / packet adaptation.
[0097]
For example, as shown in FIG. 1, B. 1, A. 2 and B. 2 (considering the features of the present invention), the conventional type that is physically close to L1 / L2 of GSM-CSD and GPRS (ie A.1 and B.1) The network has a direct control channel (or a process that works for it) that adapts L1 / L2 according to the QoS explicitly known at the peer. For example, peer A.I. 1 and B.I. 1 has access to the AT command interpreter, through which it can decide whether to run a reliable link layer protocol (RLP). However, on the other hand, 2 and B.B. 2 cannot adapt to L1 / L2 which is different from L1 / L2 for TCP flows when they started. Peer A. 2 and B.B. There is no control process (function 3) that can be applied to 2.
[0098]
GSM-CSD can only operate in a fixed mode (FEC scheme and I-mode error recovery) on circuit startup calls, even though it must send multiple flows with different QoS requirements simultaneously. is there.
[0099]
Although the invention has been described with reference to specific embodiments, the embodiments have been set forth for clarity and complete description only, and the invention is to be construed in accordance with the appended claims. The Japanese invention should not be construed as being limited to these examples. The reference signs in the claims are attached to assist the understanding of the claims and are not intended to limit the scope of the claims.
[Brief description of the drawings]
FIG. 1 is a conceptual diagram showing a communication device according to the present invention.
FIG. 2 is a flowchart illustrating an embodiment of a method according to the present invention.
FIG. 3 is a conceptual diagram showing a hierarchical structure of a protocol using TCP / IP as an example.
4 is a conceptual diagram illustrating a state in which a packet flows in a layer illustrated in FIG.
FIG. 5 is a conceptual diagram showing a concept of encapsulation in the next protocol layer of FIGS. 3 and 4;
FIG. 6 is a conceptual diagram illustrating a connection between two IP hosts via a circuit-switching GSM link and a hierarchical structure of protocols corresponding to the connection.
FIG. 7 shows the basic structure of an HDLC frame.
FIG. 8 is a block diagram of a preferred embodiment in accordance with the present invention.
FIG. 9 is a flow chart illustrating a preferred method embodiment in accordance with the present invention.
FIG. 10 is a diagram illustrating another preferred embodiment of the present invention.

Claims (28)

It is determined by the first protocol that is preset, to create a data packet having a first data structure is determined by the preset second protocol receives a data packet having a second data structure embeds each data packet having the structure of the second data packet having one or more of the first structure, based on the contents of the data packet having the structure of the second, according to a preset rule, the A device for determining a data packet having a second structure, and for adjusting an adjustable parameter of the data packet having the first data structure based on the determination result ,
The preconfigured first protocol supports at least two transmission reliability modes that can be distinguished with respect to the transmission of data packets, at least with respect to the retransmission rules of the data packets, and each created data packet has: The data packet receiving means has information on the communication reliability mode in which the data packet should be transmitted so that the data packet receiving means can determine in which communication reliability mode the data packet was transmitted. The information about the mode is inserted into each data packet by the communication device prior to transmitting the data packet,
The discriminating means (1) is configured to discriminate a data packet having the second structure according to the predetermined rule, and includes the predetermined data packet having the second structure. One or more pieces of information relating to the communication reliability mode of the data packet having the first structure are set based on a determination result. The first protocol supports two communication reliability modes, that is, a first mode including a rule for retransmitting a data packet under a predetermined condition, and a second mode in which the data packet is not retransmitted. The communication apparatus according to claim 1 . The communication apparatus according to claim 1 or 2 , wherein the communication apparatus is configured to divide a data packet having the second structure among data packets having the first structure. . The communication apparatus according to claim 1 or 2 , wherein the communication apparatus is configured to contain a data packet having the second structure in a data packet having the first structure. 5. The communication device according to claim 1 , wherein the data packet having the first structure is delivered to an output buffer (5), and the packet is sent out from the output buffer. A buffer (5) to which a data packet having the first structure is delivered and output from the data packet is sent to each data packet having the first structure according to the communication reliability mode set for each packet. The communication apparatus according to claim 1 , wherein the communication apparatus is placed in a queue. The first protocol supports two communication reliability modes, that is, a first mode including a rule for retransmitting a data packet under a predetermined condition and a second mode for not retransmitting the data packet. The communication device according to claim 6 , wherein the output buffer (5) retransmits the first mode data packet to be retransmitted with a higher priority than other data packets. Said data packet having a second structure has one or more sections containing information about the contents of the packet, said determining means (1) analyzes the one or more sections, based on the The communication device according to claim 1 , wherein the data packet having the second structure is discriminated based on the content of the data. Wherein the one or more sections are packet headers of the corresponding protocol layer, the contents of the packet communication apparatus according to claim 8, characterized in that it comprises a protocol identification information for identifying the relevant protocol. Said data packet having a second structure, a section containing information about the contents of the packet contains one or more, said determining means (1) analyzes the one or more sections, the second structure determine according to the contents of a data packet having said one or more sections is a packet header of the corresponding protocol layer, includes a protocol identification information identifying the protocol contents of the packet are associated, the packet The header has a hierarchical structure according to the protocol layer, and the discriminating means (1) for the packet to be discriminated is:
Performing a protocol identification in a header associated with the second protocol, and then comparing the protocol identifier to a rule that associates a preset protocol identifier with a preset communication reliability mode;
Communication reliability mode to be discriminated packet, if any, in the rules which the protocol identifier is stored, to set the communication reliability mode according to a preset rule, in the rules which the protocol identifier is stored If there is not, the protocol identifier in the protocol header in the layer higher by one layer in the hierarchical structure is identified, and the protocol identifier of the next protocol is stored indicating the assignment with the preset communication reliability mode. Against the rules
Where the above identification and comparison steps
Identified protocol identifier in the to be discriminated packet, assigned to one of the predetermined communication reliability mode according to one of the rules, the mode as the communication reliability mode of the packet to be the determination Set or
Corresponding to the fail-safe condition, the default value of the communication reliability mode is set in the packet to be determined, or
The communication apparatus according to claim 1 , wherein the communication apparatus repeats until any one is performed. The communication apparatus according to claim 2 , wherein the first protocol divides the packet having the second structure so as to be embedded in the packet having the first structure. 12. The communication apparatus according to claim 1, wherein the first protocol is a protocol for transmitting a packet through a link. The communication apparatus according to claim 12 , wherein the link is a wireless link. The communication apparatus according to claim 12 or 13 , wherein the communication apparatus receives a data packet having the first structure through the link. According to a first protocol which is previously set, a method for creating a data packet having a first structure,
-Based on the content of the data packet having the second structure, the received data packet having the second structure based on the preset second protocol is discriminated (S2) according to a preset rule;
-Embedding the data packet having the second structure in the data packet having the first structure (S3), and setting an adjustable parameter of the data packet having the first data structure according to the determination result Because
The preset first protocol supports at least two communication reliability modes for transmitting data packets, and the communication reliability modes can be distinguished at least with respect to data packet retransmission rules Each data packet includes information on the communication reliability mode when sending each data packet so that the data packet receiving side can specify the communication reliability mode in which the data packet was transmitted, and has a second structure. A method for creating a data packet, characterized in that the information related to the communication reliability mode of the data packet having the first structure including the data packet having the following is set by a determination result regarding the data packet having the second structure. The first protocol, the two communication reliability mode, that is, to support a first mode for determining the retransmission of the data packet under certain conditions, a second mode that does not perform retransmission of the data packet The method of claim 15 . The method according to claim 15 or 16 , wherein the data packet having the second structure is enclosed in the data packet having the first structure. The method according to claim 15 or 16 , wherein the data packet having the second structure is divided into packets having the first structure. 19. A method according to any of claims 15 to 18 , characterized in that a data packet having the first structure is passed to an output buffer (5). The data packet having the first structure is passed to the output buffer (5), said output buffer (5), the data packet having the first structure is set for each such packet, corresponding to the communication reliability mode The method of claim 15 , wherein the method writes to a queue that performs. The first protocol supports two communication reliability modes, a first mode that defines retransmission of data packets under specific conditions, and a second mode that does not perform retransmission of data packets, and the output buffer 21. The method of claim 20 , wherein (5) retransmits the first mode data packet with a higher priority than other data packets. Data packets having the second structure comprises one or more sections having information about the contents of the packet, the determining step (S2) is to analyze the one or more sections, whereby in accordance with the contents 16. The method according to claim 15 , wherein a data packet having the second structure is determined. 23. The method of claim 22 , wherein the one or more sections include protocol identification information that identifies a protocol to which a packet content of a corresponding protocol layer is associated. Said data packet having a second structure has one or more sections containing information about the contents of the packet, the determining step (S2) analyzes the one or more sections, based on this According to the content , the data packet having the second structure is determined,
The one or more sections are packet headers of corresponding protocols, including protocol identification information identifying a protocol associated with the contents of the packet;
The packet header has a layered structure according to protocol layers, and for the packet to be determined , the determination step (S2) includes:
-Performing a protocol identification in a header associated with the second protocol, and then comparing the protocol identifier to a rule that associates a preset protocol identifier with a preset communication reliability mode;
-If the communication reliability mode of the packet to be determined is in the rule in which the protocol identifier is stored, the communication reliability mode is set according to a preset rule, and the communication reliability mode of the rule in which the protocol identifier is stored If not, identify the protocol identifier in the header of the protocol that is one layer higher in the hierarchical structure, and indicate the assignment of the protocol identifier of the next protocol to the preset communication reliability mode Compare the remembered rules,
Where the above identification and comparison steps
- identification protocol identifier in the to be discriminated packet, in accordance with one of the rules, and assigned to one of the predetermined communication reliability mode, as the communication reliability mode of the packet to be the determination Set the mode or
-Set the default value of the communication reliability mode for the packet to be identified, corresponding to the fail-safe condition,
The method of claim 15 , wherein the method is repeated until either is performed. The first protocol is a packet of said second structure, in order to embed the packet of the first structure, the method according to claim 16, characterized in that the split. 26. A method according to any of claims 15 to 25 , wherein the first protocol is a protocol for transmitting packets over a link. The method of claim 26 , wherein the link is a wireless link. The communication device A method according to claim 26 or 27, characterized in that receiving a data packet having the first structure through the link.

Images